Cedars-Sinai’s new division is an AI trailblazer
The system’s Artificial Intelligence in Medicine (AIM) unit expects to serve as a resource that shares the results of its research worldwide.
One of the newest AI initiatives in healthcare, Cedars-Sinai Health System’s Artificial Intelligence in Medicine division, is designed to serve as a global resource.
“It is important to apply new technologies that benefit healthcare stakeholders at a global level,” says Sumeet Chugh, MD, director of the new division at the Beverly Hills, Calif.-based health system.
“Despite all the progress made in healthcare, there remain critical gaps in mechanisms, diagnostics and therapeutics of major human disease conditions.”
The division will present new discoveries at scientific conferences and publish results in scientific journals, Chugh says. Long-term goals include testing AI solutions in clinical trials “with eventual deployment in day-to-day clinical care.”
New division’s mission
Cedars-Sinai, an academic medical center, recently created the Artificial Intelligence in Medicine division within its Department of Medicine to “bring all current and future AI projects under one umbrella with shared resources to maximize collaboration and efficiency and to eliminate duplication of efforts,” Chugh says. “Cedars-Sinai’s AIM will also provide mentorship and education to a spectrum of stakeholders with an interest in making contributions to the field of AI in medicine.”
AI has the potential to have a huge impact on the future of healthcare, says Chugh, who also is an associate director of the Smidt Heart Institute at Cedars-Sinai.
“Despite all the progress made in healthcare, there remain critical gaps in mechanisms, diagnostics and therapeutics of major human disease conditions,” he notes. The new division is designed to help use AI to address these critical gaps, Chugh says. “AIM faculty and staff will function as AI innovators as well as custodians of patients’ healthcare interests. Using the power of AI, they will design novel tools [leveraging] the Cedars-Sinai Health System clinical data warehouse that will be ethically vetted, analyzed, validated and implemented.”
Researchers can query the data warehouse using machine learning and deep learning methods “to create novel tools and algorithms with the goal of enhancing human disease prevention, diagnostics and therapeutics,” the director of the new division explains.
Cedars-Sinai expects to collaborate with other institutions that have “high quality AI enterprises,” Chugh says. “The specific nature of these collaborations will be determined by the questions being asked, mutual interest in the tool being developed and the potential for enhancing the investigative process.”
The new division is working on projects in the areas of heart imaging, prediction of heart attacks and sudden death, and deep learning on genomics, according to Chugh. “In the very near future, we plan to expand these efforts to multiple additional areas within medicine.”
“In our AIM division, we hold an ‘AI forum’ every two weeks for open exchange of such information and learning."
Based on Cedars-Sinai’s AI research efforts so far, the organization has already learned some important lessons, the director says.
“We are learning that it is important to invest in a common infrastructure that can be used securely by multiple groups of AI investigators,” he points out. At an early stage, it is important to initiate a process by which information can be shared and discussed, he adds.
“In our AIM division, we hold an ‘AI forum’ every two weeks for open exchange of such information and learning. Once discoveries are made, it is crucial to involve stakeholders, such as physicians, early in the process to get their feedback and provide them with regular updates of progress as well as information regarding their own specific role in deployment of novel AI tools.”
Two studies published
The division already has published two studies on applying AI.
A report in the Journal of Nuclear Medicine describes using AI algorithms to identify heart attack risk in patients with already established coronary artery disease.
“Because the actual risk for recurring heart attacks differs greatly among patients, predicting future risk in patients with existing coronary artery disease can be challenging,” says Piotr Slomka, a professor of medicine who was lead author of the study. “Predicting risk, however, becomes easier and more efficient with the use of artificial intelligence.”
Another study published in the JAMA Cardiology journal describes an AI tool that can identify and distinguish between two life-threatening heart conditions – hypertrophic cardiomyopathy and cardiac amyloidosis – that are often easy to miss. “These two heart conditions are challenging for even expert cardiologists to accurately identify, and so patients often go on for years to decades before receiving a correct diagnosis,” says David Ouyang, MD, a cardiologist who was senior author of the study. “Our AI algorithm can pinpoint disease patterns that can’t be seen by the naked eye and then use these patterns to predict the right diagnosis.”